Inline @context in examples

[filip26]

Hi,
examples related to selective disclosure contain inline @context.

{
  "type": "DataIntegrityProof",
  "cryptosuite": "ecdsa-sd-2023",
  "created": "2023-08-15T23:36:38Z",
  "verificationMethod": "did:key:zDnaepBuvsQ8cpsWrVKw8fbpGpvPeNSjVPTWoq6cRqaYzBKVP#zDnaepBuvsQ8cpsWrVKw8fbpGpvPeNSjVPTWoq6cRqaYzBKVP",
  "proofPurpose": "assertionMethod",
  "@context": [
    "https://www.w3.org/ns/credentials/v2",
    {
      "@vocab": "https://windsurf.grotto-networking.com/selective#"
    }
  ]
}

Please see this discussion
w3c/vc-data-integrity#272

@vocab should not be used in a production. Having it in examples, test vectors, could lead to confusion and security issues.

(BTW an inline @context should be banned, it only adds complexity to processing that is vulnerable to issues, nothing else)

[msporny]

@vocab should not be used in a production. Having it in examples, test vectors, could lead to confusion and security issues.

Yes, agreed.

(BTW an inline @context should be banned, it only adds complexity to processing that is vulnerable to issues, nothing else)

Yes, agreed, it's unfortunate the tests include inline contexts.

We should try to remove these if possible from the test vectors and the test suites. That said, we also don't want to destabilize implementations.

/cc @BigBlueHat @PatStLouis @Wind4Greg -- what's the plan here? I suggest we switch to a driver's license, or PRC/EAD/NatCert example.

[filip26]

another option could be https://www.w3.org/ns/credentials/examples/v2. It's consistent with other VCDM2.0 based specs.

[Wind4Greg]

Hi @filip26, @msporny, and all. Currently the VC-DI-ECDSA contains three selective disclosure examples/test vectors:

1. Employment authorization document example https://w3c.github.io/vc-di-ecdsa/#representation-ecdsa-sd-2023
2. Permanent resident card example https://w3c.github.io/vc-di-ecdsa/#additional-example-for-representation-ecdsa-sd-2023
3. Fictitious windsurfing example https://w3c.github.io/vc-di-ecdsa/#enhanced-example-for-representation-ecdsa-sd-2023

At this point I'd recommend dropping the windsurfing example which uses the "incline context" stuff. It did feature arrays and nesting that tested the SD primitives a bit more. However it also confused some folks resulting in another issue to be resolved.

@filip26, @msporny, @dlongley and all let me know what you think.

[filip26]

@Wind4Greg I'm fine with any solution that removes the inline contexts. The examples you listed could be retained with the example context, which would benefit the specification. (some related examples/vectors will need to be updated accordingly)

[msporny]

+1 to removing the windsurfing example (as much as I like it)... the inline context is throwing some people.

[dlongley]

I think it's fine to remove the example that uses the inline context.

[msporny]

PR #96 has been merged to address this issue; closing.